Computing.Net > Forums > Networking > daisy-chaining switches??

daisy-chaining switches??

Reply to Message Icon

Original Message
Name: Christopher
Date: August 10, 2002 at 06:08:16 Pacific
Subject: daisy-chaining switches??
Comment:

Is ok to daisy-chain workgroup switches, would there be a perfromance decrease if (4) switches are daisy-chained throughtout a network? I have a star-bus topology that currently has (4) hubs throughout the network in four different offices. I want to replace the hubs with workgroup switches, but I keep hearing that I should setup on switch with enough ports for all of the workstations (I currenly have 16 total in four different offices). So I am asking would I see any performance increase by daisy-chaining switches in place of the hubs? In order to run (1) 16 port switch, I would have to rerun my cable, and I would rather not do that.


Report Offensive Message For Removal


Response Number 1
Name: Curt R
Date: August 10, 2002 at 06:55:05 Pacific
Reply: (edit)

Switches and ports are very similar...the big difference being that a switch intelligently routes data to only the port the target is attached to whereas a hub broadcasts to all ports at once. Switches come with uplink ports for daisy chaining...so yes, you can do it. You won't notice any real performance increase in a small LAN environment however...so if that was your reason for changing from hubs to switches...you likely don't need to bother


Report Offensive Follow Up For Removal

Response Number 2
Name: Christopher
Date: August 10, 2002 at 07:34:47 Pacific
Reply: (edit)

What if the server is running a manufacturing software database that requires alot of information to be sent to any of the workstations at any given time? Would that require that one workstation could hog up all of the bandwidth, I think switches in place of hub could alliviate some of this problem.


Report Offensive Follow Up For Removal

Response Number 3
Name: Analyst
Date: August 10, 2002 at 12:40:44 Pacific
Reply: (edit)

Without running a network monitor to measure throughput and usage it'd be hard to tell, but under those circumstances you describe, then you might be right, and see some performance increase, especially if your replacing old 10BaseT hubs with 100BaseT switches.

When daisychaining hubs/switches, there is a general 5/4/3 rule of thumb to follow. No more than 5 segments (each wire between a hub and client is a segment) connecting no more than 4 hubs stringed together, with no more than 3 of those segments populated by clients. Just remeber, when you daisychain hubs/switches, all the clients attached to the hub will be sharing bandwidth on the port of the other hub that the populated hub is attached too.

My theory is you should have as few hubs strung together or daisychained as possible, but should have at least more than one for redundency should one fail. Hope this helps and good luck.


Report Offensive Follow Up For Removal







Use following form to reply to current message:

   Name: From My Computing.Net Settings
 E-Mail: From My Computing.Net Settings

Subject: daisy-chaining switches??

Comments:

 


  Homepage URL (*): 
Homepage Title (*): 
         Image URL: 
 
Data Recovery Software




Have you ever used OpenOffice?

Yes, as my main suite.
Yes, occationally.
Yes, but only once.
No, never.


View Results

Poll Finishes In 5 Days.
Discuss in The Lounge